I apparently can't access the article at all (I can't see any links for athens/shibboleth), so I can't tell, but does the paper make it clear that the scheduling of cases is purely at random? That could have a huge effect on bias, as could other factors such as which lawyers are around for which cases excetera.

In addition of course, this is a nonrandom sample of some judges in Isreal. Generalising to assume that one will make different decisions on important cases depending on when one has eaten is probably not a good idea. That said, its probably worth thinking about at least if your decision is important, but thats actually probably more to do if one is in a good/bad mood.